魏通

魏通 Tong Wei

Ph.D., Associate Researcher

School of Computer Science and Enginnering, Southeast University

Email: weit AT seu.edu.cn

Office: Room 270, Building of Computer Science and Enginnering, Southeast University Jiulonghu Campus, Nanjing, China

I am a member of PAttern Learning and Mining (PALM) Lab

欢迎对机器学习有兴趣的同学与我一起工作！

Biography

Ph. D., 2021.12, in Department of Computer Science and Technology, Nanjing University, China; Advisor: Prof. Yu-Feng Li. 

B.S., 2016.06, in Department of Computer Science and Technology, Soochow University, China

Research Interests

I am mainly interested in machine learning (ML) and data mining (DM), especially when the data resources (size and distribution of training set, quality of labels and annotations, etc.) are limited. Deep learning (DL) with weak supervison is my current focus.

• Learning with partial, incomplete, noisy, and weak labels

• Learning with label noise: Papers [C7]

• Semi-supervised learning: Papers [C2], [J1]

• Positive and unlabeled learning: Papers [J3], [W1]

• Dealing with imbalanced & long-tailed data distribution

• Long-tailed recognition: 

• Extreme multi-label learning: Papers [C1], [C3], [C4], [C5], [J2]

• Streaming multi-label learning

• Scalable SMLL: Papers [C6]

Selected Publications 

Journal Papers

[J.3] Robust Model Selection for PU Learning under Constraint [pdf] 
Tong Wei, Hai Wang, Wei-Wei Tu, Yu-Feng Li
Science CHINA Information Science (SCIS), 2021, to appear.

[J.2] Does Tail Label Help for Large-Scale Multi-Label Learning [pdf] 
Tong Wei, Yu-Feng Li
IEEE Transactions on Neural Network and Learning Systems (TNNLS), 31(7): 2315-2324, 2020.

[J.1] Learning Safe Multi-Label Prediction for Weakly Labeled Data [pdf] 
Tong Wei, Lan-Zhe Guo, Yu-Feng Li, Wei Gao
Machine Learning (MLJ). 107(4): 703-725, 2018.

Conference Papers (* authors contribute equally)

[C.7] NGC: A Unified Framework for Learning with Open-World Noisy Data [pdf] 
Zhi-Fan Wu*, Tong Wei*, Jianwen Jiang*, Chaojie Mao, Mingqian Tang, Yu-Feng Li 
In: Proceedings of International Conference on Computer Vision (ICCV 2021, Oral), Virtual Event, 2021, pp.62-71. 

[C.6] Probabilistic Label Tree for Streaming Multi-Label Learning [pdf] 
Tong Wei*, Jiang-Xin Shi*, Yu-Feng Li 
In: Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2021), Virtual Event, 2021, pp.1801-1811. 

[C.5] Towards Robust Prediction on Tail Labels [pdf] 
Tong Wei, Wei-Wei Tu, Yu-Feng Li, Guo-Ping Yang 
In: Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2021), Virtual Event, 2021, pp.1812-1820. 

[C.4] Learning for Tail Label Data: A Label-Specific Feature Approach [pdf] 
Tong Wei, Wei-Wei Tu, Yu-Feng Li 
In: Proceedings of the 28th International Joint Conference on Artificial Intelligence (IJCAI 2019), Macau, China. 2019, pp.3842-3848. 

[C.3] Learning Compact Model for Large-Scale Multi-Label Learning [pdf] 
Tong Wei, Yu-Feng Li 
In: Proceedings of the 33rd AAAI conference on Artificial Intelligence (AAAI 2019), Honolulu, HI, 2019, pp.5385-5392. 

[C.2] Towards Automated Semi-Supervised Learning [pdf] 
Yu-Feng Li, Hai Wang, Tong Wei, Wei-Wei Tu 
In: Proceedings of the 33rd AAAI conference on Artificial Intelligence (AAAI 2019), Honolulu, HI, 2019, pp.4237-4244. 

[C.1] Does Tail Label Help for Large-Scale Multi-Label Learning [pdf] 
Tong Wei, Yu-Feng Li
In: Proceedings of the 27th International Joint Conference on Artificial Intelligence (IJCAI 2018), Stockholm, Sweden, 2018, pp.2847-2853.

Academic Services

SPC: IJCAI 2021; ACML 2021

PC: AAAI 2019, 2020, 2021; IJCAI 2019, 2020, 2022; NeurIPS 2020, 2021; ICML 2020, 2021, 2022; ICLR 2020, 2021, 2022; ECML 2020, 2021